lstmtraining query

[Samruddhi Dhake]

Hello,

lstmtraining --model_output="D:\Test\output" --continue_from="D:\Test\Dim_test.lstmf" --train_listfile="D:\Test\eng.training_files.txt"  --traineddata="D:\Test\eng\eng.traineddata" --debug_interval -1 -max_iterations 10

After running above command, I got,

Warning: given outputs 111 not equal to unicharset of 110.

Num outputs,weights in Series:

  1,36,0,1:1, 0

Num outputs,weights in Series:

  C3,3:9, 0

  Ft16:16, 160

Total weights = 160

  [C3,3Ft16]:16, 160

  Mp3,3:16, 0

  Lfys48:48, 12480

  Lfx96:96, 55680

  Lrx96:96, 74112

  Lfx256:256, 361472

  Fc110:110, 28270

Total weights = 532174

Built network:[1,36,0,1[C3,3Ft16]Mp3,3Lfys48Lfx96Lrx96Lfx256Fc110] from request [1,36,0,1 Ct3,3,16 Mp3,3 Lfys48 Lfx96 Lrx96 Lfx256 O1c111]

Training parameters:

  Debug interval = -1, weights = 0.1, learning rate = 0.001, momentum=0.5

null char=109

Loaded 2/2 lines (1-2) of document D:\Test\Dim_test.lstmf

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 0: GROUND  TRUTH : +1.5

Iteration 0: ALIGNED TRUTH : ++11..55

Iteration 0: BEST OCR TEXT : _B_f_t_t_t_t_f

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.855%, delta=27.586%, train=450%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 1: GROUND  TRUTH : +1.5

Iteration 1: ALIGNED TRUTH : ++11..55

Iteration 1: BEST OCR TEXT : _______

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.839%, delta=27.586%, train=362.5%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 2: GROUND  TRUTH : +1.5

Iteration 2: ALIGNED TRUTH : ++11..55

Iteration 2: BEST OCR TEXT : _+++++++

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.821%, delta=27.586%, train=325%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 3: GROUND  TRUTH : +1.5

Iteration 3: ALIGNED TRUTH : ++11..55

Iteration 3: BEST OCR TEXT : +++++++

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.798%, delta=25.862%, train=300%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 4: GROUND  TRUTH : +1.5

Iteration 4: ALIGNED TRUTH : ++11..55

Iteration 4: BEST OCR TEXT : +++++++

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.765%, delta=24.138%, train=285%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 5: GROUND  TRUTH : +1.5

Iteration 5: ALIGNED TRUTH : +11..55

Iteration 5: BEST OCR TEXT : ++++...

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.704%, delta=22.414%, train=266.667%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 6: GROUND  TRUTH : +1.5

Iteration 6: ALIGNED TRUTH : +1..555

Iteration 6: BEST OCR TEXT : ++.

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.519%, delta=19.704%, train=239.286%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 7: GROUND  TRUTH : +1.5

Iteration 7: ALIGNED TRUTH : +1..5555

Iteration 7: BEST OCR TEXT : +.

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.308%, delta=17.672%, train=215.625%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 8: GROUND  TRUTH : +1.5

Iteration 8: ALIGNED TRUTH : +1...555

Iteration 8: BEST OCR TEXT : .

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=5.101%, delta=16.092%, train=200%(100%), skip ratio=100%

Encoding of string failed! Failure bytes: ffffffc3 ffffff98 34 32 33 2e 31 20 30 2e 30

Can't encode transcription: '├ÿ423.1 0.0' in language ''

Iteration 9: GROUND  TRUTH : +1.5

Iteration 9: ALIGNED TRUTH : +1...55

Iteration 9: BEST OCR TEXT : .......5

File D:\Test\Dim_test.lstmf line 1 :

Mean rms=4.888%, delta=14.828%, train=200%(100%), skip ratio=100%

At iteration 10/10/20, Mean rms=4.888%, delta=14.828%, char train=200%, word train=100%, skip ratio=100%,  New worst char error = 200 wrote checkpoint.

Finished! Error rate = 100

Later, running command, 

lstmtraining --stop_training --continue_from="D:\Test\output_checkpoint" --traineddata="D:\Test\eng\eng.traineddata --model_output="D:\Test\abc.traineddata"

It am getting.

mgr_.Init(traineddata_path.c_str()):Error:Assert failed:in file ../../../../../src/training/lstmtrainer.h, line 96

My question is, does lstmtraining first generates output_checkpoints and then giving those to lstmtraining --stop_training, does it generates .trainneddata??? 

Thankyou.

Regards,

Samruddhi

[Samruddhi Dhake]

Hi,

One more question to add here is, after running 2nd command mentioned above, I am getting assert in file lstmtrainer.h, but I didn't find this file in src/training in my folder Tesseract-OCR.

Can you please me with this too?

Regards,

Samruddhi

[Samruddhi Dhake]

Hi,

Can anyone help me to resolve above issues?

Regards,

Samruddhi